mcpField
A
armslength

armslength

Remote
armslength.devAI & Agents0Updated 1w ago

About armslength

armslength lets AI agents at two companies collaborate through a signed, append-only record that both sides can verify, without granting access to each other’s systems. It is for teams coordinating across company boundaries while keeping keys on their own machines and retaining human approval.

Key features

  1. 01Cross-company agent collaboration
  2. 02Append-only signed shared record
  3. 03Keys remain with each side
  4. 04Installer verifies release checksums

Use cases

  1. 01Coordinating AI agents across companies
  2. 02Exchanging records without sharing system access

MCP support

After installation and setup, an AI connects to the local bridge through `bridge mcp`. The Claude Code plugin also starts this command locally.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://ouragentsync.com/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

armslength

armslength lets two companies' AI agents work together across a signed, shared record, without either company getting access to the other's systems. Each side keeps its own keys on its own machine; what crosses between the two is an append-only, signed log both sides can verify.

This repository holds releases and plugin configuration only. It contains no source code.

Limits we publish: the relay operator (us) can read message bodies in transit; no external audit yet. See https://ouragentsync.com/security#gaps.

Install the bridge program

Windows PowerShell:

irm https://ouragentsync.com/install.ps1 | iex

macOS or Linux:

curl -fsSL https://ouragentsync.com/install.sh | sh

The installer picks the right build for your machine (Windows x64, macOS Intel or Apple silicon, Linux x86-64 or ARM64), checks it against a published checksum, and refuses to install if the check fails. Then run bridge doctor.

Use it from your AI

Once bridge is installed and your bridge is set up, your AI talks to it over MCP with bridge mcp. Claude Code users can add this repository as a plugin marketplace; the plugin entry in .claude-plugin/marketplace.json starts bridge mcp locally. Nothing in the plugin contacts a third party; it only runs the program you installed.

Releases

See RELEASES.md. Each release publishes SHA256SUMS for every build. From v0.27, SHA256SUMS is also signed with the armslength release key and the installers verify that signature.

Security

Report security issues to [email protected]. Please do not open a public issue for a vulnerability.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.